Hello all,
As far as I understood, the time-step for pss shooting depends on "maxacfrequency" which defaults to max(pss output harmonics,10)*4*fundamental_frequency, and the error preset. This, in turn, will determine the accuracy of the subsequent periodic small signal analysis. I did the following experiment. For a linear periodically time-varying circuit, I ran pss with zero harmonics (which means maxacfreq = 40*fs, where fs is the pss beat frequency). I checked the transfer function from an input frequency "fin" to an output frequency at "fin+15*fs" by running a pac analysis and checking the transfer function around the 15th sideband. Next, I gave "maxacfrequency" to be equal to 5*fs and checked the transfer function from fin to fin+15*fs. The result was exactly identical. I expected that since the maxacfrequency is set to 5*fs, the small signal analysis will not be able to estimate the transfer function around the 15th sideband accurately. But it doesn't seem to be the case. The error preset was set to "liberal" in both the experiments.
Please help me understand what is happening here. And what is the relation between the pac sideband transfer function accuracy and the maxacfrequency (or the output harmonics set in pss).
Thanks in advance!